Fuji X-Trans Flare / Ghosting Issue

I love Fuji X-series cameras – they have exceptionally good image quality, superb handling and they are just a lot of fun to shoot with. I have completed reviewing all Fuji X cameras that I have had during the last few months, including the X-Pro1, X-E1, X-M1 and the X100S. In short, an amazing array of cameras from Fuji. One issue that I overlooked while reviewing the cameras though, was the spotted ghosting issue caused by the X-Trans sensor in rare situations, as demonstrated below (shot with the Fuji 60mm f/2.4 Macro lens).

UPDATE: this turned out to be an issue with all mirrorless cameras that have a short flange distance. Please read this post to understand the issue in detail.

Fuji X Flare Ghosting Issue

Since I am currently working on reviewing all Fuji X lenses, I had to go through each lens to test things like optical performance, chromatic aberrations, vignetting, distortion and flare / ghosting. During my flare / ghosting test, which involves photographing a scene with the sun in the frame at different apertures, I noticed a very interesting phenomenon – each lens that I used would produce spotted ghosts that looked very defined in a straight square pattern, in addition to the lens ghosts that we normally see from lenses. At first, I thought it was a lens issue. But then as I tested one lens after another, whether it was a native Fuji or a Zeiss lens, every single one of them showed the same pattern. After a couple of lenses, I realized that these patterns are not from optical characteristics of the X lenses or the types of coatings used in them, but rather internal reflections involving the X-Trans sensor.

When does it happen?

If you are wondering when you would potentially see this kind of pattern ghosting, it only happens in very rare situations, where a very bright source of light (such as sunlight) is in the frame and the lens is stopped down to f/11 and smaller apertures. At larger apertures the pattern is mixed with normal flare / ghosting and is not quite visible, but as soon as you get to f/11, the pattern shows up in most shots, depending on where the source of light is is positioned within the frame (see more on this below).

Proof that it is not lens reflections

It is normal for any lens to produce ghosting and flares when they are directed at bright sources of light. It happens due to internal reflections occurring within the lens. However, these reflections vary from one lens to another and they certainly vary by focal length. In the case of the Fuji X lenses, they each show different tolerances to ghosting and flare, but those optical characteristics are recorded separately from the pattern ghosting that I demonstrated in the very first image. Take another look at that same image and notice that the magenta / green ghosts from the lens actually show up from the center to the right of the image frame. The “rainbow” patterns surrounding the sun have nothing to do with those ghosts and those are coming from another source. Considering that the pattern is repetitive in a square box, I think that it is coming from the X-Trans sensor. Probably has to do with microlenses on the surface of the sensor that are very prone to reflections. Since the light is falling at extreme angles to the surface of the sensor when it passes through a small aperture opening, the angle of light is most likely the cause of such microlens reflections. Interestingly, this effect is minimized when the sun is in the center of the frame, as demonstrated below (shot with the Fuji 27mm f/2.8 lens at f/11):

Sunset Sample

However, the moment you move the source of light away from the center, all Fuji X-series cameras show the same problem. Here is an example of the same rainbow pattern that is seen in an image shot by the Zeiss Touit 32mm f/1.8 at the same aperture of f/11:

Zeiss Touit 32mm Ghosting Sample

And here is another example from the Fuji 18-55mm f/2.8-4 OIS shot at f/16:

Fuji 18-55mm Ghosting Sample

As you can see, this pattern appears no matter what lens is used, which shows that it has nothing to do with optical reflections.

How to reduce rainbow ghosting

Unfortunately, those of us that like to photograph sunrises and sunsets with the sun in the frame using the Fuji X series cameras will have to think of ways to reduce this problem, since there is no other way to deal with it for now. Here is a list of suggestions to reduce the effect:

  1. If it is acceptable to your framing, position the source of light as close to the middle of the frame as possible – this seems to reduce the issue. However, this might not matter if the source of light is too bright, since the pattern might actually spill all over the frame.
  2. Try not to photograph very bright sources of light in the first place. During sunrises and sunsets, the sun should not be very bright when it is close to the horizon, which will naturally reduce such reflections. See if there are any natural ways of partially blocking the sun (such as trees, branches or leaves). I have done it a few times and it actually works even at very small apertures, as demonstrated below:

    Fuji X-M1 Sample (28)

  3. Try not to go beyond f/8 if the rainbow pattern is noticeable at smaller apertures. If you are maxing out on the shutter speed, it is best to use a neutral density filter rather than stop down the lens.

Comments

  1. 1
    ) tim
    October 28, 2013 at 10:26 am

    Thanks Nasim! Okay, for a complete wrap up. Which of the x series do you suggest for the photographer who needs a back up to his D3 and and will probably grab the Fuji for quick street and travel photography?

    Out with it!

    • October 28, 2013 at 10:37 am

      Tim, get the latest Fuji X-E2 if you can. If you want to save some money, the X-E1 is also a great camera, but obviously with slower AF speed due to use of only contrast detect.

  2. 2
    ) Chris
    October 28, 2013 at 10:32 am

    I have a X100 best camera ever owned use it more than D3
    Chris

    • October 28, 2013 at 10:38 am

      Chris, I would be curious to see what the X100 does in such situations when shot directly against the sun at say f/16. Can you test it real quick and let me know? The X100 does not have an X-Trans sensor, so perhaps it does not have the same issue?

      • 9
        ) Chris
        October 28, 2013 at 10:41 am

        Will do if the Sun come out again

        • October 28, 2013 at 10:43 am

          Sounds good, thanks! I would also check your existing images and see if you can find anything shot at small apertures and the sun in the frame.

          • 16
            ) Chris
            October 28, 2013 at 11:11 am
            • October 28, 2013 at 11:16 am

              Chris, can’t really see which one you are referring to. I went though all the images and I do not see a single one with the sun in the frame and small aperture.

            • 20
              ) Chris
              October 28, 2013 at 11:26 am

              The first 2 in workshop with 500w light

            • October 28, 2013 at 11:44 am

              The bulbs are probably quite large and not as powerful as the sun. Would be great if you could do a shot with the sun in the frame.

            • 26
              ) Chris
              October 28, 2013 at 12:10 pm

              Will do

    • 38
      ) Chris
      October 29, 2013 at 5:09 am
  3. 3
    ) Miroslav
    October 28, 2013 at 10:34 am

    I reported about this problem already on dpreview xforum 3 months ago. This problem comes with every aperture and is destroying the whole image. Look close at the sun lightsource. Do you see the sensor paterns? Well, this doesnt occures on traditional bayern sensors. And thats why I have sold my x-e1 and all lenses. Love the concept of X cameras, but this is something I simply cant overlook.

    Ps:wrote fuji, gave them raw examples, no response…

    • October 28, 2013 at 10:40 am

      Miroslav, thanks for your feedback! Well, if you photograph such scenes every day then it is obviously an issue. Most people use the X-series cameras for things like travel and street photography and situations where one would encounter the above are quite rare. Not a deal breaker for me, but certainly an annoyance. I will try to reach out to Fuji to see what they say…hoping that they will respond.

      • 11
        ) Mm
        October 28, 2013 at 10:43 am

        Cool Nasim, thanks. As I said, I fell in love with the X cameras. But so many quirks! Hope they will solve this in future.

        • October 28, 2013 at 11:17 am

          Not sure if Fuji can easily solve this one to be honest – will have to see what they say.

    • 12
      ) Hans Ernst
      October 28, 2013 at 10:47 am

      I shot some 10.000 images with the XPro-1 and only very very rarely had this problem,
      definitely not something to throw the whole system out.
      Last saturday i shot with the 35mm @f2.8 straight against a sunset backlit building/trees
      All of them no problem.

      • 13
        ) Mm
        October 28, 2013 at 10:53 am

        Well, this was my everyday system. I use for work D700, so it was no problem for me. Still miss the fantastic fuji EV, manual control and weight. What I do not miss is are the raw problems and this. I made 30.000 images and every time there was a direct sunlight, I had this problem.

      • October 28, 2013 at 11:18 am

        Yup, not an issue in most situations – just when shooting at small apertures against the sun.

    • 41
      ) Christian
      October 29, 2013 at 8:34 am

      This is not a Fuji X-trans problem, the Olympus E-PL2 had it also badly, with a conventional sensor. You should have googled first, before you sell your gear. “olympus E-PL2 red dots”

      It seems to be a common problem with mirrorless camera…

  4. 5
    ) Mm
    October 28, 2013 at 10:37 am
    • October 28, 2013 at 10:41 am

      Thanks for the link! Haven’t seen that one yet, but looks like it has already been reported and discussed before. No-one in that thread seems to know the real answer, but I am pretty sure it has to do with the X pattern and microlenses on the sensor – definitely not lenses.

  5. 14
    ) Doug K.
    October 28, 2013 at 11:01 am

    This looks more like a moire-type issue to me, possibly a side-effect of their unique anti-moire algorithms that are required with their unique sensor layout and lack of an AA filter (?). I don’t see how this can be caused by internal reflections from the sensor/micro-lenses.

    That’s just a totally uninformed theory… If so, then it might be possible for a firmware update to reduce this effect.

    • October 28, 2013 at 11:33 am

      Doug, can’t be moire, since moire is only visible on very fine patterns of texture – this is a very different problem. The X-Trans sensor has a different pattern than Bayer and that could be one of the reasons. However, note that this only happens at very small apertures and note how the ghosts are angled as you get further away from the light source – I am very inclined to say that this is from microlenses rather than the pattern itself.

      • 27
        ) Doug K.
        October 28, 2013 at 12:23 pm

        I dunno. It looks like a beat/interference pattern to me. It certainly could be due to microlenses, giving the characteristic varying angle appearance. I was using the term “Moire” very loosely; I really should have said an interference pattern. I’m just not sure it is due to reflections. To be more specific, I was hypothesizing that it is due to an interaction of their non-Bayer demosaicing algorithm, which might have unintended consequences due to their fancy anti-Moire filtering. That might cause an interference pattern, and it could easily be heightened by a very small aperture.

        I’m guessing that it is NOTHING like conventional flare.

      • 33
        ) theNeverKings
        October 28, 2013 at 7:57 pm

        Hi Nasim.

        Due to the regularity of the pattern and the manner in which the light is being refracted/separated into individual wavelengths, I suspect (as mentioned by Doug K) we are looking at an interference pattern caused by texture (CFA?) on the surface of the sensor. Strangely enough I also notice a very similar effect when observing the reflections of the downlights on my ceiling being reflected on my iPad. The effect is not as strong but it is present.

        In general, any surface that has been polished/ground will carry the marks of that process, even though they may be invisible to the naked eye. If the process is a regular (as opposed to random) process interference can observed under the appropriate conditions.

        At least it’s not too unattractive, even reminding me of some of the effects found in an old COKIN filter pamphlet I have lying around somewhere.

        Cheers.

        BTW: Thanks for all the hard work you and your staff put into PhotographyLife!

  6. 15
    ) Kevin
    October 28, 2013 at 11:09 am

    Hi Nasim
    Very interesting. I bought a Fuji X20 earlier this year as a walk about and back up for my D600
    ( which I was having problems with ! ) . I have not encountered any such problems and have been so impressed with the Xtrans system I was thinking of up grading to the XE 2. The only issue I have with the X 20 is the lack of wide angle – I’m used to a 16-35 on my D600! Now I’ve got a picture taken with the D600 and 16-35mm lens which looks to me a bit like one of your shots above and yes it was at f16. Anyway I can send it to you for comment ? email ?
    Thanks

    • October 28, 2013 at 11:37 am

      Kevin, you can send the image through the submit content form on the top of the page. Is the sun very bright in the frame, or partially blocked like above?

      • 29
        ) Kevin
        October 28, 2013 at 1:13 pm

        OK -Tried it – not sure it worked as I got a screen that said Comments closed- I sent it anyway.
        The Sun is Blocked as I hope you can see.

        • October 28, 2013 at 1:20 pm

          Kevin, I looked at the image and that’s a very different issue – the flare and ghosts you see on your D600 is caused by the lens. There are no repetitive patterns as shown in the above images…

          • 36
            ) Kevin
            October 29, 2013 at 3:32 am

            Thanks Nasim- I see what you mean about repetitive patterns. I’ve just re read your review of the 16-35mm Lens and you didn’t find any harsh flare or ghosting. Do you think there might be a problem with my lens ?

            • October 29, 2013 at 3:54 pm

              No, I don’t think you have a problem. Sometimes it is just the angle of the sun that can create flare/ghosting even in lenses that have excellent resistance to it. Also, filters can potentially add more flares/ghosts, so make sure that you use a high quality filter.

  7. October 28, 2013 at 11:31 am

    I actually like it, if it is not easily provoked. Can already see a creative use for it. :) But I am pretty sure Fujifilm will fix it in future X-Trans iterations with some sort of microlens coating or something.

  8. 28
    ) Roar Arne Velle
    October 28, 2013 at 1:07 pm

    I believe it is a side effect of small aperture and very short flange distance, which in turn gives the big light angles.

    • October 28, 2013 at 1:18 pm

      That is another possibility! Will have to test my OM-D E-M5 to see if it has the same problem or not. Cloudy today, will have to wait for the sun to show up.

      • 43
        ) Marco
        October 29, 2013 at 9:37 am

        Thanks Nasim! I have a EM-5 and I’ve never seen a behavior like this. Definitely curios to see your test results (today it’s raining here so cannot try for myself)

  9. 32
    ) peter nadort
    October 28, 2013 at 5:08 pm

    hello Nasim, after this great test what will be the following? Drop down?

    wish you luck with that.

    regards, peter

  10. 34
    ) KnightPhoto
    October 28, 2013 at 8:28 pm

    Reminds me of the old Cokin start filters! ;-)

    Good advice in the article Nasim…

    • 35
      ) KnightPhoto
      October 28, 2013 at 8:29 pm

      *star not *start

  11. 37
    ) Kevin
    October 29, 2013 at 4:50 am

    Hi Nasim. Have you tried the same type of shot with the X-A1? This has a non-X-Trans sensor so perhaps it should be ok? That may prove your thesis around the sensor type.

  12. 39
    ) Christian
    October 29, 2013 at 7:29 am

    Hi Nasim,

    thanks for the read. You seem to think, this is a X-trans sensor or a Fuji problem…

    It´s not! This is a common problem with lots of cameras – my Olympus Pen E-PL2 had it in the same way, I think even my good old Panasonic L1 had it. Can´t recall , if that´s a problem only with mirrorless cameras, but you´ll find posts about this in every camera forum, with lots of makers.

    Seems to be normal in digital photography.

  13. 40
    ) Christian
    October 29, 2013 at 7:35 am

    It´s called “red dots” and sometimes “red spots” in the Olympus forums – see p.e. here:

    http://www.flickr.com/groups/e-pl2/discuss/72157625883894721/

    and here, further down the page:

    http://visualsciencelab.blogspot.de/2011/01/olympus-epl2-final-installment-kirks.html

    I think, the conclusion was, that there are red dots in every mirrorless camera to some extends.

    • 42
      ) DrBob
      October 29, 2013 at 9:33 am

      I don’t think so. That one was a EPL-2 fault (and was even less intense) and was due to bad sensor. No other cameras has this problem ASAIK. If Fuji has the same behavior is indeed a sensor fault

      • 44
        ) Christian
        October 29, 2013 at 9:55 am

        I recall other cameras had this too? Sigma DP1 ect…

        So it is obviously no X-trans related, but a general sensor problem that can be found in various cameras.

    • October 29, 2013 at 3:55 pm

      Christian, thanks for the links. Indeed – this seems to be a common problem on mirrorless cameras. I will shoot with my E-M5 when the sun comes out and update the article. I would be curious to know what exactly causes such issues…

  14. 45
    ) Christian
    October 29, 2013 at 9:57 am
  15. 46
    ) Christian
    October 29, 2013 at 10:00 am

    I´m not saying, this isn´t a problem.

    It is, but Olympus and Panasonic cameras suffer from this too…

    I think lots of people think it´s normal lens flare and don´t see it as a problem, as it only happens at rare circumstances.

    • October 29, 2013 at 3:26 pm

      Christian, thanks for bringing this up – it was certainly something new to me as well. I will test my Olympus OM-D E-M5 later this week when the sun comes out and will update the article.

      • 50
        ) marco
        November 13, 2013 at 1:32 pm

        any news about EM5 results?

Comment Policy: Although our team at Photography Life encourages all readers to actively participate in discussions, we reserve the right to delete / modify any content that does not comply with our Code of Conduct, or do not meet the high editorial standards of the published material.

Leave a Comment